Add LoRA training scripts and fix bake-off token budget

- training/scripts/train_lora.py: Unsloth QLoRA trainer for qwen3:8b
- training/scripts/train_lora.sh: Launch script for steel141 RTX 3090 Ti
- eval/bakeoff.py: Fixed token budget (400->1500) that caused qwen3
  models to exhaust tokens on thinking, added --no-think flag
- agent/serve.py: Default model changed to gemma3n:e4b

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

agent/serve.py

@@ -37,7 +37,7 @@ from agent.prompts.system_prompts import get_prompt
 
 DEFAULT_CONFIG = {
     'ollama_url': 'http://192.168.0.179:11434',
-    'model': 'qwen3-coder:30b',
+    'model': 'gemma3n:e4b',
     'rcon_host': '127.0.0.1',
     'rcon_port': 25577,
     'rcon_password': 'REDACTED_RCON',